A new copyright fight is putting Anthropic’s handling of music works under fresh scrutiny. A cohort of music publishers led by Concord Music Group and Universal Music Group is suing the AI company, alleging that it illegally downloaded more than 20,000 copyrighted songs, including sheet music, song lyrics, and musical compositions.
The publishers said in a statement on Wednesday that damages could amount to more than $3 billion. According to the statement, that would make the dispute one of the largest non-class action copyright cases filed in U.S. history.
What the publishers are alleging
The case centers on how Anthropic allegedly obtained music-related copyrighted works. The publishers say the company illegally downloaded more than 20,000 copyrighted songs, a group of works that includes sheet music, song lyrics, and musical compositions.
That distinction matters because the dispute is not framed only around whether AI models can be trained on copyrighted material. The publishers are also pressing the claim that Anthropic acquired the material through piracy.
The lawsuit directly attacks the company’s public positioning. It says:
“While Anthropic misleadingly claims to be an AI ‘safety and research’ company, its record of illegal torrenting of copyrighted works makes clear that its multibillion-dollar business empire has in fact been built on piracy,”
Anthropic did not respond to TechCrunch’s request for comment.
Why the Bartz v. Anthropic case matters here
The lawsuit was filed by the same legal team from the Bartz v. Anthropic case. In that earlier case, a group of fiction and nonfiction authors accused Anthropic of using their copyrighted works to train products like Claude.
Judge William Alsup ruled in Bartz v. Anthropic that it is legal for Anthropic to train its models on copyrighted content. But the ruling also drew a line around how that content was obtained: he pointed out that it was not legal for Anthropic to acquire that content via piracy.
That split is central to the new music publishers’ case. The earlier ruling did not treat AI training and allegedly pirated acquisition as the same issue. Instead, it separated the legality of model training from the legality of getting the works in the first place.
The Bartz v. Anthropic case became a $1.5 billion outcome for Anthropic. Impacted writers received about $3,000 per work for roughly 500,000 copyrighted works. The source article notes that while $1.5 billion appears significant, Anthropic was valued at $183 billion.
How the music case expanded
The music publishers had already filed a lawsuit against Anthropic over its use of about 500 copyrighted works. But the new action is broader because of what the publishers say they learned through discovery in the Bartz case.
According to the publishers, that process revealed that Anthropic had also illegally downloaded thousands more works. They then tried to amend their original lawsuit to include the piracy issue.
The court denied that motion back in October. The ruling said the publishers had failed to investigate the piracy claims earlier. That denial led the publishers to file a separate lawsuit instead.
The separate lawsuit also names Anthropic CEO Dario Amodei and co-founder Benjamin Mann as defendants. That makes the new case not only a dispute between publishers and an AI company, but also one that reaches named company leaders.
The stakes for AI and copyrighted music
The case arrives at a moment when AI companies face continued pressure over the data behind their products. The source article identifies Claude as one of the products involved in the earlier authors’ case, and the new music lawsuit focuses on the alleged downloading of copyrighted songs and related materials.
For music publishers, the alleged scope is substantial. More than 20,000 copyrighted songs are at issue, and the claimed damages could reach more than $3 billion. The works described in the complaint cover different parts of music publishing, including lyrics, sheet music, and compositions.
The legal conflict also shows how copyright disputes around AI can turn on several separate questions:
- whether copyrighted content was used to train AI products;
- how that content was obtained;
- whether the works were downloaded illegally;
- what damages should apply if the claims succeed.
The publishers’ case is built around the acquisition issue as much as the use issue. That focus follows the logic highlighted in Bartz v. Anthropic: training on copyrighted content and acquiring copyrighted content through piracy are treated as different legal questions in the source article’s account.
For Anthropic, the lawsuit adds another major copyright challenge after Bartz v. Anthropic. For publishers, it is an attempt to turn alleged discovery findings into a separate case focused on music works, damages, and the company’s alleged torrenting of copyrighted material.